Are Imacon scans supposed to be good, or....?

[Mammoth_research]

I just had my local lab process some of my first 4x5 E6 film I've ever shot. A couple of the positives came out great, so I asked for them to be scanned on their imacon scanner (not sure on model specifics). The files are something like 42 megabytes for just two scans (seems really small). They honestly look really disappointing in terms of resolution. They completely fall apart at 200% on photoshop and it looks almost worse than some 35mm I've seen scanned.

What am I missing here? I've heard nothing but good things about imacon/hasselblad quality, although I know real drums scans are far superior...

Should I just pay for pricey drum scans? I want to print at least 16x20

Thanks

 

[koraks]

fall apart at 200%

Well...that's one pixel of image data for every 4 pixels on your monitor. I wouldn't expect any magic.

Is there a possibility of uploading sections of these scans that demonstrate the problem?

 

[warden]

Yeah 42mb from a 4x5 sounds small. What are the pixel dimensions of the image? They may need to try again at a higher resolution.

 

[MattKing]

A historical question: Did the practice of characterizing scan quality in terms of file size alone arise in the days of dial-up modems?
It is so unhelpful!

 

[Steven Lee]

Resolution is not measured in megabytes. Everything falls apart at 200% in Photoshop. @Mammoth_research you need to give us more to troubleshoot. Why not simply share a link to the original scan you got from the lab?

 

[Mammoth_research]

thanks everyone - I do understand the sarcasm lol. I know what I sound like right now and I hate myself for even asking this question, being that I'm the type of person who's head spins when talking anything photoshop or digital in general.

That being said:

There isn't really anything wrong with the scans, nor is there anything to troubleshoot. There is however, an obvious lack of resolution when you consider the extremely fine detail of a 4x5 sheet of Provia.

All I am looking for here, I guess, is validation or consolation in my experience with this Imacon scanner. Has anyone here scanned 4x5 on an Imacon and also been unhappy with the results? What did you seek out to solve this problem?

Thanks

 

[Mammoth_research]

I am not really sure how to upload an image for you all since the file limit here on photrio is 2mb... maybe you all know (I have like zero experience with digital image stuff and am bad with computers)

 

[koraks]

There is however, an obvious lack of resolution when you consider the extremely fine detail of a 4x5 sheet of Provia.

OK, I see what you mean. What are the pixel dimensions of the scans you got? I.e. how many pixels on the long side, how many on the short side? And/or did they tell you at what dpi they scanned?

I am not really sure how to upload an image for you all since the file limit here on photrio is 2mb

You could e.g. upload one full image downsized to e.g. 1Mb to give an impression of the entire frame, and then a couple of cropped areas that show the full resolution. That way you can make do with a few files (2-3), each of them being small enough to fit the upload limit.

 

[dokko]

the top end Imacon/Hasselblad scanners have a 8000pixel line scan sensor, which means they can scan 4x5" up to 2000ppi. this will result in a bit over 300MB in 16bit RGB or a bit over 150MB in 8bit.
so if your 42MB scan is 8bit, that would equal to around 1000ppi scan resolution.

for a good 16x20" print I would aim for at least 1500ppi, which shouldn't be too hard on a good scanner (once you need over 4000ppi it gets tricky).

however, the scan resolution is only part of the scan quality. there's also dynamic range, noise, and color accuracy.

if you have a dropbox or google drive you could upload the original and share a link. or if you want I can send you an upload link in a direct message and host it for a couple of weeks.

 

[Mammoth_research]

dokko and koraks thank you for your super helpful answers! Those figures are very helpful.

So with those figures, and assuming the unit they used was top of the line at 8000 pixel line scan, it would not be unreasonable to ask the lab for a higher quality scan. I wonder why they don't use the machines max potential when making a "Premium Scan" (in their words)?

dokko, thank you for the offer to host the file, but I will first try a google drive link of my own and see if that works out.

 

[Mammoth_research]

Here is a link, please let me know if it works and is visible.

BIG CLARIFICATION HERE: The 42mb file size was two images combined. This single image (one out of two that were scanned) is only 21 megabytes!! So even smaller than you calculated, dokko

 

[dokko]

So with those figures, and assuming the unit they used was top of the line at 8000 pixel line scan, it would not be unreasonable to ask the lab for a higher quality scan. I wonder why they don't use the machines max potential when making a "Premium Scan" (in their words)?

with a 848/949 they would be able to scan around twice the resolution of what you have now, with a 4 times larger file size.

if I were you I'd simply ask which scanner thy use exactly.

but it might be best to look at you current scan first.

edit: I see you already uploaded. I sent a request but you can also change the option to download possible for everybody.

 

[Mammoth_research]

I accepted your request and changed the permissions to anyone with a link can view!

 

[MattKing]

thanks everyone - I do understand the sarcasm lol.

Sorry - no sarcasm meant toward you.
I was actually expressing frustration about that when labs/scanning services use that method to indirectly hint at scan resolution.

 

[dokko]

got it, thanks.
looking at the file, it's 3024 pixels wide including the rebate, so it would just make for an okish 10x8" print.

that said, the colors look washed out, some of the highlights are clipped, and it's severely over sharpened, so I wouldn't call this a premium scan and would be hesitant to use a lab which delivers scans like that for demanding work.

If you are experienced with digital images you could look for a lab that offers affordable 3F scans (basically RAW scans) at maximum resolution and adjust the colors yourself using the flexcolor software (or simply photoshop for reversal film).

If you'd rather that the lab does it for you, than I would look for a lab who has the skills and good communication.

 

[Mammoth_research]

Sorry - no sarcasm meant toward you.
I was actually expressing frustration about that when labs/scanning services use that method to indirectly hint at scan resolution.

Gotcha yeah it has been particularly tricky for me to understand how to get what I'm looking for from the digital scanning world because of thing like this. I have such high hopes because of work from people like Ben Horne and Michael Strickland who use this large format to digitally scanned to pigment print workflow

 

[Mammoth_research]

got it, thanks.
looking at the file, it's 3024 pixels wide including the rebate, so it would just make for an okish 10x8" print.

that said, the colors look washed out, some of the highlights are clipped, and it's severely over sharpened, so I wouldn't call this a premium scan and would be hesitant to use a lab which delivers scans like that for demanding work.

If you arer experienced with digital images you could look for a lab that offers affordable 3F scans at maximum resolution and adjust the colors yourself using the flexcolor software.

If you'd rather that the lab does it for you, than I would look for a lab who have the skills and good communication.

Yes I saw all of these issues too and to be honest, I wouldn't be surprised if some are my fault, especially a bit of over exposure. However, it does seem like I should be getting more out of my film - especially since what I saw on the light table was a lot nicer than this.

I am not experienced with digital images and frankly I don't know what any of that means lol. But I am willing to learn and I do have access to photoshop as a learning tool.

I think the next step is asking the lab if they think they can realistically do better than this. Then, if not, I will seek out some other lab to work with.

 

[Focomatter]

There is the possibility that the computer connected to the Imacon was not up to the task at full rez. Imacons use obsolete connectors so maybe an old and under-powered but still functioning computer was used.

 

[Mammoth_research]

There is the possibility that the computer connected to the Imacon was not up to the task at full rez. Imacons use obsolete connectors so maybe an old and under-powered but still functioning computer was used.

Gotcha, that is an interesting thing that I wouldn't have thought of. I will ask them if that is a possibility

 

[dokko]

There is the possibility that the computer connected to the Imacon was not up to the task at full rez. Imacons use obsolete connectors so maybe an old and under-powered but still functioning computer was used.

unlikely, I used an Imacon 848 connected to a macmini from around 2010 for years and could easily scan 400MB files.

 

[warden]

Here is a link, please let me know if it works and is visible.

BIG CLARIFICATION HERE: The 42mb file size was two images combined. This single image (one out of two that were scanned) is only 21 megabytes!! So even smaller than you calculated, dokko

Is there any chance that the service provider sent you a preview image for your approval instead of the final output? Honestly that scan is smaller than I would expect from a 35 mm negative on a home scanner, and your image was from a vastly larger format of film. With a scan of 4 x 5 film you should be drowning in detail.

 

[Mammoth_research]

Is there any chance that the service provider sent you a preview image for your approval instead of the final output? Honestly that scan is smaller than I would expect from a 35 mm negative on a home scanner, and your image was from a vastly larger format of film. With a scan of 4 x 5 film you should be drowning in detail.

I am pretty sure this is the final output... I will have to get on the phone with them tomorrow

 

[Mammoth_research]

dokko - I was just checking out your scanning service website and WOW you are the real deal!! Thanks for helping out the newbies even at your level of expertise. Super cool stuff

 

[koraks]

anyone with a link can view!

I've had a quick peek; @dokko of course gave an expert assessment, which should be taken note of.

I feel there may be place for some nuance. Firstly, the washed out (I'd call them 'neutral' or 'unsaturated') colors may be a deliberate choice in order to leave the issue of saturation to the client. There's sufficient color information to take this pretty much any way you want. We could argue about whether this is the most fortunate choice in terms of s/n ratio, but it seems to me there's ample flexibility in terms of the color data that's there.

Secondly, considerable areas do indeed look blown out (which in fact emphasizes the low saturation). It turns out that there's a little color information in most of these areas. However, because the scanning people decided to bunch this all together in the right side of the histogram, trying to create more differentiation in these areas results in significant posterization. See e.g. here:

Note also in the second snippet, there's hard clipping especially in the red and green channels. Now, you also said this:

I wouldn't be surprised if some are my fault, especially a bit of over exposure

That's of course never a fortunate decision when it comes to exposing slides. I can't tell what's in the actual slide, but it's conceivable that there wasn't all that much to work with as a result of partial overexposure. Nevertheless, I would expect better color resolution in the highlight areas even in a moderately overexposed slide. The expansion of the tonal scale of the file so that it occupies the entire histogram has been in done in such a way that I find it hard to believe it would optimally exploit the capabilities of the scanning hardware. I think @dokko can confirm this based on his experience with these machines.

As to resolution and sharpening - no contest; it's a low-res scan and indeed, it has been 'spiced up' considerably to that it looks "nice and sharp", but this limit your own choices in controlling final output. As it is, the file is indeed insufficient for a big print. Resolution-wise, an 8x10" print from this file will look quite good - if not for the color problems.

It's not a very good scan by any account.

Overall, a scan from a 'prosumer' grade flatbed (e.g. Epson V-series) will generally be significantly better as long as it's operated by someone who has a bit of an idea of what they're doing. Indeed, you can make a scan from a 4x5" transparency on something like an Epson V700 that will print perfectly fine at 16x20".

 

[dokko]

Thanks for helping out the newbies even at your level of expertise. Super cool stuff

thanks for the kind words.
as you might have seen in the technical section of my page, I'm not a huge fan of the Imacon scanners, but for a 16x20 print of a 4x5" slide it can provide excellent results if operated properly.

koraks already pointed out that even a good consumer flatbed can do well for this print size, it really comes down to experienced operators which care about their work. once you go over 20x25" the scanner model starts to become important.

my experience with lab services is: either they do care and have some pride in their own work or not.
if they don't care, talking to them is a pretty futile endeavour since you always have to complain to get them to deliver ok results, but they will never be outstanding anyway.

clipped highlights in any scan for me always is a tell sign that it's sloppy work because it's dead easy to avoid.