comparison between Epson V700 and Howtek D4000

[zhengjdc]

I did a comparison test a few month ago.

This particular film is captured with EOS 3 body + 24mm 1.4L lens, this picture is captured on manfrotto tripod (very steady). Film used is 135 Provia 100F 36 exposures. Venue: Shanghai, China.

The D4000 had a loosen belt back then, so my image of this drum scanner had a bit fuzziness around the edge, but I don't have the time to redo the test after getting new belt. Scanned at 4000DPI @ 6 micron with Silverfast Ai.

v700 is done with Doug's wet mounting kit after 1 whole day of careful adjustment of heights at 3.7mm. Scanned at 6400DPI with Epsonscan.

Both were wetmounted with Aztek SMF mounting fluid. v700 scanning results were sharpened moderately, drum scanning results were left as is.


v700:
http://farm4.static.flickr.com/3483/5794038283_e694de7fff.jpg

D4000:
http://farm4.static.flickr.com/3154/5794594672_0e59b3f3f5.jpg

v700:
http://farm3.static.flickr.com/2063/5794596542_c3fa16cde7_b.jpg

D4000:
http://farm4.static.flickr.com/3407/5794036679_25497c26b6.jpg


I supposed after belt replaced, D4000 should be showing more clarity. Also, after tried Vuescan for v700, the result is better than the epsonscan software, especially with proper exposure control, but unfortunately, I don't have time/energy to redo a more specifiy comparison.

Both costs almost the same: V700 $500 used on kijiji with doug wet mounting kit $120(incl shipping). D4000 $700 used comes with 4 drums and mounting station.

-----------------------------------------------------------------

2024 update:

Sony A7R IVA, Signa 105mm MACRO, captured at f5.6 with cinelite light table (CRI >99%, R9 >95%)

246660

246661

246662

246663

246664

246665

[zhengjdc]

More comparison:

v700:
http://farm4.static.flickr.com/3650/5794595184_61bfd505aa.jpg

D4000:
http://farm4.static.flickr.com/3616/5794595264_3f93e45c1c.jpg

v700:
http://farm3.static.flickr.com/2114/5794594862_23107675b0.jpg

D4000:
http://farm6.static.flickr.com/5193/5794035979_c5f6bf3d0d_m.jpg

[zhengjdc]

v700:
http://farm4.static.flickr.com/3409/5794035603_ee2c0f02cd.jpg

D4000:
http://farm3.static.flickr.com/2642/5794594576_b7552de72a_m.jpg

v700:
http://farm4.static.flickr.com/3068/5794035799_6134632057.jpg

D4000:
http://farm6.static.flickr.com/5062/5794035377_90a7b8d3e9_m.jpg

[Brian Ellis]

Thanks for posting the results of this test, I always appreciate it when other people take the time to do this kind of thing because I hate doing it myself. But I think you could have saved yourself the testing time. I don't think anyone has ever claimed that the Epson flatbeds are any good with 35mm film except maybe for a really tiny snapshot or for the web. But for a real print from 35mm film, even just an 8x10, they're pretty much useless. I didn't find mine real good even for moderate sized prints from 6x7 but some people think they're o.k. for that, certainly not 35mm though.

I guess I should mention that I'm not sure it's appropriate to post test results from 35mm film in a large format photography forum. Doesn't bother me but it might bother some people.

[zhengjdc]

Thanks for posting the results of this test, I always appreciate it when other people take the time to do this kind of thing because I hate doing it myself. But I think you could have saved yourself the testing time. I don't think anyone has ever claimed that the Epson flatbeds are any good with 35mm film except maybe for a really tiny snapshot or for the web. But for a real print from 35mm film, even just an 8x10, they're pretty much useless. I didn't find mine real good even for moderate sized prints from 6x7 but some people think they're o.k. for that, certainly not 35mm though.

I guess I should mention that I'm not sure it's appropriate to post test results from 35mm film in a large format photography forum. Doesn't bother me but it might bother some people.

The purpose is to demonstrate scanners, not a discussion of films. The best way to test a scanner is by using 135 format with superb lens as most LF film has much much lower PPI which even a $50 crappy scanner delivers somewhat OK result.

Most people found v700 sux becoz they don't do a proper height adjustment, or not as carefully as I did. V700 still lacks shadow details pe se but other than that, I found it being pretty good at scanning films, even 135, and vuescan+doug's wet mounting kit is a must have.

however, since nowadays, a decent working drum scanner costs no greater than $1000 put v700 really into a tough position as the wet mounting on v700 is such a pain in the a$$, I still remembered that I spent great deal of time to roll out an air bubble trapped there.

[sanking]

however, since nowadays, a decent working drum scanner costs no greater than $1000 put v700 really into a tough position as the wet mounting on v700 is such a pain in the a$$, I still remembered that I spent great deal of time to roll out an air bubble trapped there.

It is true that one can find drum scanners for sale at bargin prices. However, it is important to note that most of this equipment is nearly two decades old, very large and heavy, and one can expect to spend a fair amount of time with upkeep to keep them going. Same is true of the professional flatbeds.

Odd that you found wet mounting with the Epson a pain in the a$$. My own experience is that fluid mounting with a flatbed scanner is quite a bit easier than with a drum.

Sandy

[zhengjdc]

It is true that one can find drum scanners for sale at bargin prices. However, it is important to note that most of this equipment is nearly two decades old, very large and heavy, and one can expect to spend a fair amount of time with upkeep to keep them going. Same is true of the professional flatbeds.

Odd that you found wet mounting with the Epson a pain in the a$$. My own experience is that fluid mounting with a flatbed scanner is quite a bit easier than with a drum.

Sandy

The roller that I had is too soft to remove some air bubbles and it's always a pain to remove them all....

on the other hand, mounting station has great rubber roller, great clamping power, most likely 1 roll is enough to remove bubbles.

after I got the drum scanner, v700 is left there collecting dusts. :D

[Brian Ellis]

The purpose is to demonstrate scanners, not a discussion of films. The best way to test a scanner is by using 135 format with superb lens as most LF film has much much lower PPI which even a $50 crappy scanner delivers somewhat OK result.

Most people found v700 sux becoz they don't do a proper height adjustment, or not as carefully as I did. V700 still lacks shadow details pe se but other than that, I found it being pretty good at scanning films, even 135, and vuescan+doug's wet mounting kit is a must have.

however, since nowadays, a decent working drum scanner costs no greater than $1000 put v700 really into a tough position as the wet mounting on v700 is such a pain in the a$$, I still remembered that I spent great deal of time to roll out an air bubble trapped there.

I understand that you were comparing scanners. But making the comparison by using a film format that one of the scanners is unsuited for, and then getting a result that anyone familiar with Epson flatbed scanners could have predicted without doing any testing, seems pointless to me. But it's your time and if you think it's a meaningful test that's fine.

[zhengjdc]

forgot to insert the original picture, here we go:

http://farm3.static.flickr.com/2646/5809157412_5707b08719_b.jpg

circled out with red rectangles were where comparison took place.

[archivue]

you are comparing sharpness... but looking at the entire image, colors are off...
i have a V750 myself... just to make preview... using C41 negative film, when i'm in the process to make a large print or a book, i'm renting an imacon X1 or X5... the main reason is color !

[Lenny Eiger]

you are comparing sharpness... but looking at the entire image, colors are off...
i have a V750 myself... just to make preview... using C41 negative film, when i'm in the process to make a large print or a book, i'm renting an imacon X1 or X5... the main reason is color !

If you want more color, use a drum scanner... PMT's are far more sensitive than CCD. However, if you are noticing colors that ought to be more blue or more yellow, red or green, or color casts, this is not the way to judge a scanner. Those things are almost always some distance off based on scanner operator choices and are easily fixed in PhotoShop.

Lenny

[Ash]

My V700 is terrible for colour and transitions. It takes forever to carefully adjust using a variety of profiles to try and get the best tonal range, then correct the colours to something normal. Still, some shots are bad, some are good. Some don't make sense.

As far as sharpness I'm not sure that the two comparison shots are fair if you sharpened one and not the other?

[SergeyT]

The major issue with this test is running 4000dpi vs. 6400dpi

[Lenny Eiger]

The major issue with this test is running 4000dpi vs. 6400dpi

It's way worse than that. This test (and probably this thread) should be discarded entirely. The PMT's in this machine are bad and the drum speeds are off on the scanner. He posted this on the scan Hi End list... It doesn't represent what a properly tuned drum scanner can do....

I think its fine for a beginner to say, "Hey, look what I found." However, both pieces of equipment should be working properly and tuned, and the operator should be properly trained before anyone takes any of this seriously.

Lenny

[zhengjdc]

It's way worse than that. This test (and probably this thread) should be discarded entirely. The PMT's in this machine are bad and the drum speeds are off on the scanner. He posted this on the scan Hi End list... It doesn't represent what a properly tuned drum scanner can do....

I think its fine for a beginner to say, "Hey, look what I found." However, both pieces of equipment should be working properly and tuned, and the operator should be properly trained before anyone takes any of this seriously.

Lenny

I am curious how can you tell the drum speed is off and the PMT is bad?

[Lenny Eiger]

I am curious how can you tell the drum speed is off and the PMT is bad?

I am also on the Scan Hi-End list and you posted it there. You are having more than your share of problems, and I feel for you.... but it also sort of invalidates the test a little....

I'm not trying to give you a hard time. I just think your comparisons will be better when you get all the issues worked out....

Lenny

[8x10 user]

An odd statement. The amount of color in a scan is mostly due to the settings on the scanner. It really doesn't have anything to do with PMT or CCD. Your statement seems a little contradictory to me. Also a more sensitive scanner would have a higher D-max which if over the D-max of the film would produce washed out, "lower color" images; unless the operator corrects for this by adjusting the curve or black point.

Again, I am still not sure about the CCD versus PMT arguments that you make. I've made C41 Scans on a Supreme and I was blown away by the amount of color definition and subtle tonality the machine was able to pull out of the short D-range of a color Neg.

A CCD scanner like the supreme measures way more actual photons then a PMT drum scanner. The amount of light that strikes the CCD is likely to be more then 10,000 times what reaches a PMT in a drum scanner.

Of course some CCD scanners uses smaller and less sensitive sensors (Supreme = better then epson scanner). The actively cooled CCD sensor in a Supreme is one step better in terms of tonality noise and D-range then the one used by betterlight. This is because of the active cooling which brings the sensor below ambient temperatures. At really cold temperatures a CCD's noise response can be as low as one miscounted photon per hour. Of course the supremes CCD is not that cold, but it is very respectable.

I would think that a supreme would do better with subtle tonal changes than many PMT scanners due to being a true 16 bit machine. A 10, 12, or even 14 bit scanner would be a little more limited due to the electronics in the machine.







If you want more color, use a drum scanner... PMT's are far more sensitive than CCD. However, if you are noticing colors that ought to be more blue or more yellow, red or green, or color casts, this is not the way to judge a scanner. Those things are almost always some distance off based on scanner operator choices and are easily fixed in PhotoShop.

Lenny

[Lenny Eiger]

An odd statement. The amount of color in a scan is mostly due to the settings on the scanner. It really doesn't have anything to do with PMT or CCD.

Sure it does. It's a different technology, and it has a higher sensitivity. All you have to do is look at real scans....

Lenny

[8x10 user]

Again, I have owned and used 3 different high end CCD and 2 drum scanners. I have seen results first hand and have made my own comparison tests, on kodachromes, E100g and color neg. The Aztek scanner is limited to a D-max of 3.88 and is beat in shadow performance and D-max by a Eversmart Supreme that is in good condition.

For color negs that doesn't matter of course because the D-max isn't that high. I like drum scanners for color negs because of the aperture control. But in terms of being able to pull tones off of a color neg the supremes are amazing. Of course I used the highest quality modes with multi-sampling all done in a 16 bit environment. The drum scanners that I have used are also good, however I have seen scans from lesser drum scanners and I don't think a howtek 4000 or 4500 would be as good for color negs as a Supreme that is in good condition.

Demo one sometime at a respectable dealer. Just make sure you learn about the software first and that you have the descreening options turned off.

Is it true that the aztek is not a real 16 bit scanner? I know it is a rebranded and upgraded howtek 8000. If you look on the manual on their site for the 8000 it clearly states that the analog to digital convertor is only 12 bits.

This site shows that the Supreme is a pure 16-bit machine.

http://www.promarketinc.com/catalog/scanners/IQ-Scanner-vs-Eversmart.pdf

The light sensitivity might be higher with a PMT but each of the 8000 CCD spots on a supreme receives light at the same time, they are pretty large and receive a lot of light. Plus it can scan multiple times and average the results for more pure data. A PMT converts light to electricity and than multiples the current in a vacuum. A CCD takes light, converts it to energy, builds a charge for each pixel location than sends it to the electronics to be converted into bits and bytes. In the supreme it really doesn't doesn't need to multiply the current because it receives more light. The reason why a PMT uses is a vacuum is to prevent heat from effecting the multiplication. A supremes CCD is actively cooled to deal with heat issues. The Peltier is able to make things really cold. The same technology is used for small portable freezers and is also used by NASA for their high end telescopes. Peltier coolers on CCD sensors are rare, the only other scanner that uses one is the imacon 949/ X5. It costs $8,000 for a new sensor for the supreme.

The parts that convert the charges to bits and bytes (the electronic single handling) are all top of the line in a supreme. They wanted to make the best CCD scanner that scanner money could buy so they only used the best parts they could get. Unlike smaller companies they had the funds to have any type of custom component produced just for their company.

Some CCD scanners deserve more respect then others and the Eversmart Supreme deserves more than you give it.

[Lenny Eiger]

The Aztek scanner is limited to a D-max of 3.88 and is beat in shadow performance and D-max by a Eversmart Supreme that is in good condition.

We have a problem here with numbers. The Aztek specs quoted are from a test made that was intended to be an independent comparison. They were supposed to be real numbers. The other scanners aren't using real numbers, they are using theoretical ones. We had a rousing argument about this a couple of years ago on Scan Hi-End. The Imacon was quoting some amazing numbers, and one participant finally came up with how those numbers were calculated. They were theoretical, and not real. The Scannerforum test was actual more accurate. Certain manufacturers threatened to sue and made them take the numbers off the site...

I would put the Premier's DMax capabilities at the top of the list. They use PMT's and have tuned their electronics to a higher level of performance than any other scanner company todate. This is partially because their development efforts were later, after some of the companies were already hurting. If you look at the comparisons you will see that there is a category for "Quality CCD's" which have DMax values much lower than any of the drums. A CCD is not going to match a PMT unless it is poorly tuned.... there are physics involved, which can't be denied...

The Premier is a full 16 bit machine....



This site shows that the Supreme is a pure 16-bit machine.
http://www.promarketinc.com/catalog/scanners/IQ-Scanner-vs-Eversmart.pdf


This is a site where used equipment is sold. They are not scanner experts, by any measure......


Some CCD scanners deserve more respect then others and the Eversmart Supreme deserves more than you give it.

Supreme's are impressive scanners. They did some very interesting things. But a PMT is a PMT..

If you're happy with your scans, great. That's all that matters....

Lenny

[Tyler Boley]

I'm not sure what's being discussed here any more or how much of this is helpful to the topic beyond turf protection. At one point there was mention of "off" color problems with scanning color negs. This is often difficult, and generally a software and/or workflow problem. Other than completely bizarre and unusual color negs, somehow exposed beyond all reason and or cooked to death in processing, the density ranges are well withing the capabilities of all but the lowest end scanners utilizing any of the common technologies. From the mention of "off" color, there was the suggestion that "more" color was needed. What is "more color" in terms of scanning? As mentioned, color negs are not challenging to capture in terms of the film density range (all of the color?), from then on it's a matter of how that data is defined... color management and editing. Per channel capture range ability may be relevant to transparency scanning due to potential density range limitations, and "more color" could be related to per channel level count I suppose, but I doubt that's what going on here.
When discussing comparisons of various hardware and software we are all using, and technical abilities, some kind of more precise language would be helpful. Some of us are supposed to know what we a talking about.
Tyler

[8x10 user]

I'm sorry but this is simply not true. I have no idea how the numbers on the scanner forum were calculated but I have done my own tests with Kodachrome and the supreme pulled more shadow detail... To say that an Eversmart Supreme has a d-max of 3.0 is fraudulent. That is almost 1/10th of the sensitivity required to reach the black point of Kodachrome film. Who knows why Phil gave that value and how it was determined but its just not correct. If you fluid mount, use the newer software, have a fresh bulb in it, have the calibration done correctly, are using a scanner that is in good shape, and have the settings done correctly, then the D-max is much closer to 4.0, if not higher.

You have been saying these things about other scanners for years. You come up with "logical" reasons why you say your scanner is the best but most of the "logic" is full of holes and is clearly marketing hype to me. I'm not saying your scanner sucks but you make a lot of false statements to try to convince people who don't know better, that it is the best thing since sliced bread. It's not the best scanner in world, by every criteria, for every purpose, without any equal. You are trying to over sell your service by putting down other equipment in untrue ways. In my opinion, it is bad business.

Look at the scanner comparison test on this forum, the best scanners for shadow noise performance are the Supreme, the Crosfield, and the ICG. I am let down by the general sharpness on those Supreme scans but it is not my scanner and I did not make the scans... I suspect there was an error in focusing caused by not fluid mounting or it focused on a particle on top of the mylar.

So anyway I am calling you out. Put your money where your mouth is.... How much do you want to wager that the supreme is limited to a D-max of 3.0? We can test this... Buy a couple of these. I'm sending my scanner out for service soon but when it comes back I will do some fluid mounted test scans and show it is indeed higher then 3.0 and much closer to 4.0.

http://www.edmundoptics.com/onlinecatalog/displayproduct.cfm?productID=2743

P.S. I have a lot of respect for the kind people at Aztek and I am sorry that I constantly have to debunk myths that are spread about their equipment in comparison to other scanners.




The Scannerforum test was actual more accurate. Certain manufacturers threatened to sue and made them take the numbers off the site...

I would put the Premier's DMax capabilities at the top of the list. They use PMT's and have tuned their electronics to a higher level of performance than any other scanner company todate. This is partially because their development efforts were later, after some of the companies were already hurting. If you look at the comparisons you will see that there is a category for "Quality CCD's" which have DMax values much lower than any of the drums. A CCD is not going to match a PMT unless it is poorly tuned.... there are physics involved, which can't be denied...

[Peter De Smidt]

Theoretical superiority and superiority in practice don't necessarily go together, especially if the theory isn't very good.

For example, consider the sensor size argument. At one time, it was argued that making ccd and cmos sensors with smaller photo sites, i.e. ones with higher resolution but the same light sensing area, would lead to higher noise and less dynamic range. It's physics, don't ya know! Well, current cameras put the lie to that, as they have smaller sensors but less noise and greater dynamic range. For instance, compare the results from a Nikon D200 and D7000. The D7000 has higher resolution, lower noise and greater dynamic range even though it has smaller photo sites.

[Lenny Eiger]

You have been saying these things about other scanners for years. You come up with "logical" reasons why you say your scanner is the best but most of the "logic" is full of holes and is clearly marketing hype to me. I'm not saying your scanner sucks but you make a lot of false statements to try to convince people who don't know better, that it is the best thing since sliced bread. It's not the best scanner in world, by every criteria, for every purpose, without any equal. You are trying to over sell your service by putting down other equipment in untrue ways. In my opinion, it is bad business.

I have no interest in putting down anyone's else's scanners. Many of my comments are directed at the difference between consumer level scanners and professional level ones. I still believe there is a big difference between CCD technology and PMT, and its physics....


look at the scanner comparison test on this forum

I find these quite faulty. The post-processing makes many of the examples invalid. I find it very hard to read anything from the comparisons.


So anyway I am calling you out. Put your money where your mouth is.... How much do you want to wager that the supreme is limited to a D-max of 3.0? We can test this... Buy a couple of these. I'm sending my scanner out for service soon but when it comes back I will do some fluid mounted test scans and show it is indeed higher then 3.0 and much closer to 4.0.

Sorry, not interested. I don't want to go and spend over $100 to satisfy your curiosity. I get much of my information from the folks at Aztek, who have tested all this stuff and their word is good enough for me... I already have a scanner, and there is no reason for them to lie to me so I'll buy another... if you want to go and look up the difference in response between a CCD and PMT on some web site, then by all means do so. Leave me out of it...

I do think the Premier is the top, maybe an ICG can get there as well. But more often that not its meaningless. The main thing is the operator, of course. I see a lot of folks getting great scans from all kinds of equipment. I don't know why this conversation is even worth this amount of typing.....

Lenny

[zhengjdc]

Did another scanning comparison with my Sony A7R IVA with Sigma 105mm MACRO lens. See updates.

IMHO, the MACRO lens beats both drum scanner and V700 and I utilized a special dry mount technique allowing 100% evenness thus can be taken at sharpest aperture at F5.6.

[rawitz]

This is only valid for 135 film. If you scan 120 film or sheet film, the winner changes rapidly to the flatbed scanner.