Played with the Epson 750 yesterday

[Jack Flesher]

My dealer got their first shipment in yesterday. I happened to be there as they unpacked it and thought I'd offer my initial impressions.

I have a 4990 and was interested in seeing how much better -- resolution and DR -- the 750 is over the 4990. To clarify, this is not a complete test, only my observations on scanning a single 4x5 test tranny of a typical landscape image on one sample of the 750 scanner. The image is of Bryce, with lots of detail, bright clouds, dark shadows -- and of course, dust.

I scanned at 2400 DPI, sharpening and ICE both off, but the levels slider adjusted and optimised for the tranny. I used the Epson driver as the store does not install Silverfast on any of their machines since it creates conflicts with some of the other scanner drivers.

First thing I noted is the prescan time is notably slower than the 4990's -- not a huge deal, just noticed it. The scan itself takes about the same time on either machine.

Now to the juice.

First thing was the 750 scan was soft. Not even the dust was sharp. I took the feet off (minus position, lowering the holder) and re-scanned. This was even worse, so I added the feet back in to the plus location, raising the holder. The result was essentially the same as the normal position scan, so I am unimpressed with the resolution at this point -- it most certainly is not better than the 4990 and appears to be slightly worse.

But there is some good news: D-max, or dynamic range, whatever you want to call it, is notably better on the 750. I can only quantify it by saying it looks like about half a stop gain in the shadows which may not be earth-shattering but IMO is noteworthy.

Other comments: The holders are only improved in that you can adjust the feet, but it appears the adjustment does not do a lot, so no real gain here. The wet-mount adapter is nothing to get excited about either. It is a glass tray that sits on top of the regular scanning glass, so you scan now through two sheets of glass. You can wet mount up to about a 5x7 neg, but not much more so your dream for an inexpensive 8x10 wet scanner has been dashed. I did not try a wet mount on the 4x5, so perhaps that will render a better resolution scan -- we'll have to wait for others to report their findings on that. I did learn the wet-scan tray is available as a stand-alone purchase and can be used on the 700, though it is unclear if it will work on the 4990. I doubt it since the 700 series appears to use a series of small hole patterns to tell the scanner which holder is being used.

That's it for now -- I am planning to go back this afternoon to see if their scanner tech has been able to improve things. I'll report any updates.

Me? For now I remain unimpressed and will stick with my 4990 -- I'd rather have the added resolution than the extra DR.

Cheers,

[Scott Rosenberg]

thanks for the first impressions, jack. i think the photo-i fella found that a minor adjustment to the height of the holders sharpened up the scans.

scott

[Ted Harris]

thanks for the first impressions, jack. i think the photo-i fella found that a minor adjustment to the height of the holders sharpened up the scans.

scott
We found that, even with minor adjustments the improvement didn't bring the resolution up to that of the 4870/4990 unless you were scanning at the 6400 dpi setting.

As for DMax, I'd be interested in seeing some one-to-one samples using exactly the same settings as our tests using the Stouffer Step Wedge showed the DMax of the 700 was exactly the same as that of the 4990. The 750 could be slightly better as a result of the anti reflection coatings.

[Bruce Ho]

Scott, Jack did make what height adjustments you can, taking the feet off, then putting them in at the "+" position. I didn't see any improvement either from the standard position.

The worst thing to me so far with the 750 is that the 120 holder doesn't work at all, at least on my 6x7 stuff. In an attempt to hold the film firmly, the holder actually grabs it as you lock it in a way that pushes the sides together and forces an unusable bulge in the film. After many tries at getting it to behave, I gave up and am using the holder from my 4870, which the 750 scanner recognizes ok, and which I have to shim up to an acceptable focus height. Not pretty, but it works.

Re the softness of the 750 scan, my conjecture at this point - and I offer it only as that and without wanting to get stuck as a 750 advocate and defender - is that the 750 is doing less sharpening (even with sharpening de-selected) than the 4870. Out the chute, my 4870 scan does look crisper, but when I work closely with output from both scanners I'm not sure the 4870's scan is truly sharper. The 750's scan - and I can't demonstrate this quantitatively (although I'm sure there are lots of you out there with more expertise than I have who could prove or disprove it) - seems to be cleaner, more accurate, and "fuller" in some way. After I'm done with whatever manipulations in Photoshop, cleaning up as possible with Neat Image, and then doing a final sharpening (I use smart sharpen in PS), I'm almost always more pleased with the overall result with the 750 than with the 4870.

[Jack Flesher]

We found that, even with minor adjustments the improvement didn't bring the resolution up to that of the 4870/4990 unless you were scanning at the 6400 dpi setting.

As for DMax, I'd be interested in seeing some one-to-one samples using exactly the same settings as our tests using the Stouffer Step Wedge showed the DMax of the 700 was exactly the same as that of the 4990. The 750 could be slightly better as a result of the anti reflection coatings.

Interesting comments Ted! I did not measure the difference in shadows, just compared them side to side -- nor did I use the exact same settings, but rather optimised the settings for each scanner using the levels adjust. Both got a slight gamma boost and I could have easily given the 750 a slightly larger boost on the gamma. Again this was an informal comparison and honestly, after I saw the lack of resolution on the first scan I was not overly careful with those adjustment sliders. I serioulsy doubt anti-reflective coatings on the scan lens would add the half stop or so I saw.

[Michael Gordon]

I'm curious why everyone gets excited about each Epson scanner revision. I still own the 2450, and as has been proven by many, each version since has only offered marginal improvements and hardly in the way of resolution or dynamic range.

You get what you pay for. A sub-$800 scanner may cut it for some, but Epson scans just aren't competitive enough (I use mine for web scans only). Why do Epson scanners command so much attention from this community?

[Doug Dolde]

Why do Epson scanners command so much attention from this community?

We're cheap bastards looking for a bargain.

[Ted Harris]

You get what you pay for. A sub-$800 scanner may cut it for some, but Epson scans just aren't competitive enough (I use mine for web scans only). Why do Epson scanners command so much attention from this community?

Michael, I think they command so much attention because, aside from the Microtek 1800f which is now discontinued, there aren't any choices without making a quantum leap in price and even then you are talking purchasing used until you get into the $10,000 range. OTOH, I totally agree (although I seldom say it) that you do get what you pay for and expecting a scanner that sells for under $800 with software to perform on the same level as a scanner whose software alone costs several times $800 is not a sensible expectation. An even better comparison .... I paid about the same to have my Screen Cezanne crated and shipped as the purchase price of a V700.

Jack, one more comment on your scanning times. As you havce read in my earlier posts ont he V700, your experience mirrors mine exactly. I have now run all the timetests on the machine that it seems sensible to run and the bottom line is that scans just take a long time. I posed the question directly to Epson as one of a number of questiosn I wanted answered as part of my review of the machine and they never answered the specific question to my satisfaction. The July-August "View Camera" with the review should be on the street next week.

[MJSfoto1956]

every test I have ever done with an Epson indicates that they overstate the "resolution" by a factor of 2x. The simple fact is this: there is no way the cheap lens they use can actually "resolve" 6400 dpi. I would guestimate that 3200 dpi is the maximum effective real data you will get out of this machine. The previous 4800dpi Epsons were equally overstated in that 2400dpi was their "real" resolving power.

Now that being said, 3200dpi (or 2400 for that matter) x 4"x5" = a damn lot of pixels.

I only wish to point out the error of continuing to promote "6400dpi" as a somewhat bogus claim by Epson's marketing gurus. We all should know better...

J Michael Sullivan

[Jack Flesher]

The simple fact is this: there is no way the cheap lens they use can actually "resolve" 6400 dpi. I would guestimate that 3200 dpi is the maximum effective real data you will get out of this machine. The previous 4800dpi Epsons were equally overstated in that 2400dpi was their "real" resolving power.


I'm not convinced it was ever as high as 2400 -- I think it has been 1600 since the 3200, and they just increased the multiplier to 3x for the 4800/4990. The V700/750 may be 3200 x 2, but I would not be surprised if it were just 1600 x 4...

[Tim Lookingbill]

Just curious but how are you all calculating real resolution in terms of spi for the Epson? Is it on edge sharpness or the amount of crisp detail?

I get expected edge sharpness from 35mm scans at 4800spi off my refurbed Epson 4870 that if you sharpened it enough could still look good for an enlarged print at 300 dpi. I don't understand how you claim Epsons don't have the spi stated.

[Doug Fisher]

>>I did learn the wet-scan tray is available as a stand-alone purchase and can be used on the 700, though it is unclear if it will work on the 4990.<<

Unless your 4990's optimum film suspension height for "best" focus is way off, I really don't think the fluid tray will produce superior results on a 4990. The 4990's optimum suspension height is supposed to be right at 1 mm. FWIW, most people who have tested and reported back to me in regard to their 4870 and 4990 scanners have found their optimum height to be around the 1.8 mm range. After testing with my digital caliper, it looks like my 750's fluid tray suspends at close to 3.5 mm. If your fluid tray is the same (because low cost glass thickness tolerance can vary quite a bit and will affect things), I don't think you would be getting the best results.

>>I think it has been 1600 since the 3200<<

No doubt the real resolution is probably half or less and we could debate for a while what the final numbers are/were, but in terms of sharpness, I definitely saw a good improvement from the 2450 to the 3200. In my experience the improvment from the 3200 to the 4870 was much smaller (but certainly there) and the difference from the 4870 to the 4990 was nada. If someone already owns a 4870 or 4990, I really encourage them to try shimming their holders. There are other advantages to the V series, but you might be surprised at the improvement you get with your current scanner and how well you shimmed scans compare to a V series. You can do this fairly inexpensively on your own by utilizing cut squares of plastic shim stock or even Mylar. You might find that an hour of your time and $10 worth of material convinces you that you don't need to spend $450+ on an upgrade.

I agree with others here who think the Epsons are still a decent "value" proposition - except that at their higher prices and current performance level, probably not as much of a value as in the past.

Doug
---
http://www.betterscanning.com

[joolsb]

This is an interesting discussion about softness but I never really expected much in the way of sharpness from my v700. All that matters to me is that it cleans up nicely when USM is applied - which it does.

What I find more worrying is how easily banding appears on scans from the V700. I've been scanning some Velvia 50 5x4s recently and this film seems particularly poor in this respect. It seems that the scanner has a hard job dealing with gradual transitions from bright to very dark areas. Has anyone else noticed this and is there anything I can do about it (apart from doing my best to disguise it in Photoshop)?