A question for Tamron 17-70mm owners

[jm10]

I am in the process of evaluating a new Tamron 17-70mm lens. Among other things I noticed that a X-T2 camera does not report correct focal distances between 17 and 24 mm - it is always 17 mm. Starting with 24 mm and beyond it works fine. Has anyone encountered this?

Another problem (a more serious one) is sharpness. Evaluated the lens both with a test chart and in the field. My 10 year old 18-55 mm Fujifilm lens consistently comes out on top (I do have a good copy!). Most of the reviewers report opposite results. I have attached a couple of comparison shots; the word "mushy" comes to mind when examining the Tamron shots...have to pixel-peep here at 100% to see the difference. Pictures are unprocessed JPEGs out of camera

All pictures were taken with a sturdy tripod using two X-T2s with Fuji and Tamron lenses. Swapping cameras does not help.

Does it look like a bad copy?

jacob




1 - 18-55mm




2 - 17-70 mm




3 - 18-55 mm




4 - 17-70mm

 

[Erik Baumgartner]

The Fuji looks clearly superior to me (you do have a nice copy), however the Fuji likely has significant "Lens Optimizer" processing helping it out that the Tamron may not. You might want to compare RAW files as well.

Also, try f/5.6, you should get better detail out of both of them at wider apertures, and with less diffraction reduction intervention processing muddying the waters with the jpegs.

I don't have the Tamron (I have the Fuji 16-55), but from what I've seen it probably ought to be capable of better than what we're seeing here.

 

[jm10]

The Fuji looks clearly superior to me (you do have a nice copy), however the Fuji likely has significant "Lens Optimizer" processing helping it out that the Tamron may not. You might want to compare RAW files as well.

Also, try f/5.6, you should get better detail out of both of them at wider apertures, and with less diffraction reduction intervention processing muddying the waters with the jpegs.

I don't have the Tamron (I have the Fuji 16-55), but from what I've seen it probably ought to be capable of better than what we're seeing here.

Thanks for responding Erik!

You are probably right about the "Lens Optimizer" for Fuji JPEGs. I did compare RAW files for one of the shots. The difference is significantly less noticeable but I still prefer the Fuji shot. There is no sharpening applied in LR, only an equal measure of Shadows and Whites adjustments to match the JPEGs.

I used f/10 for a reason. Most of my landscapes shots are within the f/8 to f/11 range; if I had to use f/5.6 to get decent resolution for this lens, my interest in the 17-70 mm evaporates rather quickly...

But I did notice one peculiarity while using the test chart. The resolution numbers obtained from f/4.0 to about f/10 are almost the same; the Excel charts are very flat for most of the focal range for this lens. This appears to be very abnormal. Most of the lenses I evaluated using the same chart peak around f/5.6 (as you noted) but this one does not. I repeated the measurement 3 times with the same result.

jacob




Fujifilm - from RAW - no sharpening




Tamron - from RAW - no sharpening

 

[Erik Baumgartner]

The Fuji looks clearly superior to me (you do have a nice copy), however the Fuji likely has significant "Lens Optimizer" processing helping it out that the Tamron may not. You might want to compare RAW files as well.

Also, try f/5.6, you should get better detail out of both of them at wider apertures, and with less diffraction reduction intervention processing muddying the waters with the jpegs.

I don't have the Tamron (I have the Fuji 16-55), but from what I've seen it probably ought to be capable of better than what we're seeing here.

Thanks for responding Erik!

You are probably right about the "Lens Optimizer" for Fuji JPEGs. I did compare RAW files for one of the shots. The difference is significantly less noticeable but I still prefer the Fuji shot. There is no sharpening applied in LR, only an equal measure of Shadows and Whites adjustments to match the JPEGs.

I used f/10 for a reason. Most of my landscapes shots are within the f/8 to f/11 range; if I had to use f/5.6 to get decent resolution for this lens, my interest in the 17-70 mm evaporates rather quickly...

But I did notice one peculiarity while using the test chart. The resolution numbers obtained from f/4.0 to about f/10 are almost the same; the Excel charts are very flat for most of the focal range for this lens. This appears to be very abnormal. Most of the lenses I evaluated using the same chart peak around f/5.6 (as you noted) but this one does not. I repeated the measurement 3 times with the same result.

jacob


Fujifilm - from RAW - no sharpening


Tamron - from RAW - no sharpening

Really very close here, I suspect your Tamron is probably not "bad". I'd still try a wider aperture on a subject that requires less DOF (wide open too, which is important if you're paying for that constant f/2.8). At wider focal lengths you can generally manage plenty of DOF and great sharpness at wider apertures if you focus carefully. I'm not especially familiar with the two lenses that you're testing, but my similar 16-55 is significantly sharper at f/5.6 than at f/10, and can manage quite a bit of DOF (and sharpness across the frame) even wide open at f/2.8 at the wide end of its range. I would hope the Tamron could do it too.

 

[jm10]

The Fuji looks clearly superior to me (you do have a nice copy), however the Fuji likely has significant "Lens Optimizer" processing helping it out that the Tamron may not. You might want to compare RAW files as well.

Also, try f/5.6, you should get better detail out of both of them at wider apertures, and with less diffraction reduction intervention processing muddying the waters with the jpegs.

I don't have the Tamron (I have the Fuji 16-55), but from what I've seen it probably ought to be capable of better than what we're seeing here.

Thanks for responding Erik!

You are probably right about the "Lens Optimizer" for Fuji JPEGs. I did compare RAW files for one of the shots. The difference is significantly less noticeable but I still prefer the Fuji shot. There is no sharpening applied in LR, only an equal measure of Shadows and Whites adjustments to match the JPEGs.

I used f/10 for a reason. Most of my landscapes shots are within the f/8 to f/11 range; if I had to use f/5.6 to get decent resolution for this lens, my interest in the 17-70 mm evaporates rather quickly...

But I did notice one peculiarity while using the test chart. The resolution numbers obtained from f/4.0 to about f/10 are almost the same; the Excel charts are very flat for most of the focal range for this lens. This appears to be very abnormal. Most of the lenses I evaluated using the same chart peak around f/5.6 (as you noted) but this one does not. I repeated the measurement 3 times with the same result.

jacob


Fujifilm - from RAW - no sharpening


Tamron - from RAW - no sharpening

Really very close here, I suspect your Tamron is probably not "bad". I'd still try a wider aperture on a subject that requires less DOF (wide open too, which is important if you're paying for that constant f/2.8). At wider focal lengths you can generally manage plenty of DOF and great sharpness at wider apertures if you focus carefully. I'm not especially familiar with the two lenses that you're testing, but my similar 16-55 is significantly sharper at f/5.6 than at f/10, and can manage quite a bit of DOF (and sharpness across the frame) even wide open at f/2.8 at the wide end of its range. I would hope the Tamron could do it too.

Very true - I am paying for the constant f/2.8 which I don't really need; would be perfectly happy with a f/4. Have you noticed how many professional landscape photographers shoot at f/16 with their FF Sony and Nikon cameras? That would be around f/11 APS-C...quite a bit into diffraction territory which is evidently not a primary criteria in selecting the aperture setting. I was drawn to this lens with the promise of a better sharpness and extended focal range on the long side.

It also raises another point - there are quite a few non-Fuifilm lenses around these days. Does it mean that their owners cannot get a full advantage of their lenses while taking JPEG images and have to process RAW images to get their full potential? Just thinking out loud - you don't have to answer this

 

[Erik Baumgartner]

The Fuji looks clearly superior to me (you do have a nice copy), however the Fuji likely has significant "Lens Optimizer" processing helping it out that the Tamron may not. You might want to compare RAW files as well.

Also, try f/5.6, you should get better detail out of both of them at wider apertures, and with less diffraction reduction intervention processing muddying the waters with the jpegs.

I don't have the Tamron (I have the Fuji 16-55), but from what I've seen it probably ought to be capable of better than what we're seeing here.

Thanks for responding Erik!

You are probably right about the "Lens Optimizer" for Fuji JPEGs. I did compare RAW files for one of the shots. The difference is significantly less noticeable but I still prefer the Fuji shot. There is no sharpening applied in LR, only an equal measure of Shadows and Whites adjustments to match the JPEGs.

I used f/10 for a reason. Most of my landscapes shots are within the f/8 to f/11 range; if I had to use f/5.6 to get decent resolution for this lens, my interest in the 17-70 mm evaporates rather quickly...

But I did notice one peculiarity while using the test chart. The resolution numbers obtained from f/4.0 to about f/10 are almost the same; the Excel charts are very flat for most of the focal range for this lens. This appears to be very abnormal. Most of the lenses I evaluated using the same chart peak around f/5.6 (as you noted) but this one does not. I repeated the measurement 3 times with the same result.

jacob


Fujifilm - from RAW - no sharpening


Tamron - from RAW - no sharpening

Really very close here, I suspect your Tamron is probably not "bad". I'd still try a wider aperture on a subject that requires less DOF (wide open too, which is important if you're paying for that constant f/2.8). At wider focal lengths you can generally manage plenty of DOF and great sharpness at wider apertures if you focus carefully. I'm not especially familiar with the two lenses that you're testing, but my similar 16-55 is significantly sharper at f/5.6 than at f/10, and can manage quite a bit of DOF (and sharpness across the frame) even wide open at f/2.8 at the wide end of its range. I would hope the Tamron could do it too.

Very true - I am paying for the constant f/2.8 which I don't really need; would be perfectly happy with a f/4. Have you noticed how many professional landscape photographers shoot at f/16 with their FF Sony and Nikon cameras? That would be around f/11 APS-C...quite a bit into diffraction territory which is evidently not a primary criteria in selecting the aperture setting. I was drawn to this lens with the promise of a better sharpness and extended focal range on the long side.

I guess everybody's different, I shoot my 16-55 at f/2.8 all the time in low light and almost never go much beyond f/9 or f/10.

It also raises another point - there are quite a few non-Fuifilm lenses around these days. Does it mean that their owners cannot get a full advantage of their lenses while taking JPEG images and have to process RAW images to get their full potential? Just thinking out loud - you don't have to answer this

No, not really. I think that 18-55 receives considerably more in-camera "optimizing" than just about any other lens (you seem to have one of the exceptionally good ones, but most are quite mediocre, IMO). I only shoot RAW myself, but I think most third party lenses do fine as SOOC jpegs with just the basic geometry and vignetting correction they get in-camera (the ones that do, anyway).

 

[Peter Foiles]

I recently compared my 17-70 with a rental 16-55 on my X-H2 and found them to be comparable except at the long end (55+) where I consistently preferred the 17-70. I have used the 18-55 in the past but found the 16-55 better at the edges and it was my most used lens at the time. I shoot raw exclusively and did the comparison at f5.6 which is near optimum for both lenses.

A comment on dof. I also shoot Sony FF and avoid apertures beyond f11 and for apsc beyond f8. The Fuji X landscape photographer Andy Mumford I believe avoids going beyond f11 for his work.

Based on my experience, the 17-70 is at least as good as the 16-55 if not better and should be better than the 18-55 though I have not made that direct comparison.

 

[Lettermanian]

I would be concerned that there could be a communication problem if you're not seeing all the available focal lengths. My 17-70 displays all fl's on my X-T3. Do you have any other newer body to test it with, in case the X-T2 has an issue? Also, did you take test shots with OIS and VC on vs off, to see if the respective lens IS systems are suspect? I don't pixel peep but I can say that I'm happy with the sharpness of my 17-70, though I don't have an 18-55 to compare it with. I've had four 18-55's over my years with Fuji, and I would say two were good copies and two were mediocre. Looks like yours is a very good one.

 

[jm10]

I recently compared my 17-70 with a rental 16-55 on my X-H2 and found them to be comparable except at the long end (55+) where I consistently preferred the 17-70. I have used the 18-55 in the past but found the 16-55 better at the edges and it was my most used lens at the time. I shoot raw exclusively and did the comparison at f5.6 which is near optimum for both lenses.

A comment on dof. I also shoot Sony FF and avoid apertures beyond f11 and for apsc beyond f8. The Fuji X landscape photographer Andy Mumford I believe avoids going beyond f11 for his work.

Based on my experience, the 17-70 is at least as good as the 16-55 if not better and should be better than the 18-55 though I have not made that direct comparison.

Thanks for sharing your experience Peter. It does help.

I have not seriously considered the 16-55 mm lens because of its weight and OIS. But the 17-70 reviews appeared quite promising. One of my expectations was sharpness improvement over my 18-55 and so far I fail to see this. Is this due to sample variations? I don't know.

I understand your dof comments and concerns. From my test results using the resolution charts both the 18-55 and 17-70 behave in a similar manner. At f/10 the effects of diffraction are almost unnoticeable (at least to my eye). Starting with f/11 I do see some resolution loss due to diffraction. Although this slight degradation can be recovered in pp. What is strange with the 17-70 the resolution seems to be "flat" from f/4 to f/10. The typical peak around f/5.6 is missing. This is based on JPEG image evaluation. Erik commented earlier that the 17-70 JPEGs may not benefit from Fujifilm's "optimization" which is applied to native lenses. I am not sure I want to evaluate the performance of a lens based on RAW image results...

 

[jm10]

I would be concerned that there could be a communication problem if you're not seeing all the available focal lengths. My 17-70 displays all fl's on my X-T3. Do you have any other newer body to test it with, in case the X-T2 has an issue? Also, did you take test shots with OIS and VC on vs off, to see if the respective lens IS systems are suspect? I don't pixel peep but I can say that I'm happy with the sharpness of my 17-70, though I don't have an 18-55 to compare it with. I've had four 18-55's over my years with Fuji, and I would say two were good copies and two were mediocre. Looks like yours is a very good one.

Thank you for responding Lettermanian.

Unfortunately I don't have a newer body. Both X-T2s behave the same. I could probably live with this If the image quality was meeting my expectations, but it does not. Yes, tried the OIS on and off - makes no difference. Tamron does claim that the lens senses the presence of a tripod - I guess the feature does work.

I heard many stories about the 18-55mm copy variability and feel very lucky to have a very good copy. So should it set the high bar for this new 17-70 $800 lens?

I just looked at my XC 50-230 test results obtained quite a few years ago and at 55 mm it easily beats the the 17-70 mm lens (using the same resolution chart and the same camera). Does not feel right to me...

 

[sluggy_warrior]

I am in the process of evaluating a new Tamron 17-70mm lens. Among other things I noticed that a X-T2 camera does not report correct focal distances between 17 and 24 mm - it is always 17 mm. Starting with 24 mm and beyond it works fine. Has anyone encountered this?

mine also goes from 17mm to 20mm, no 18mm or 19mm

Another problem (a more serious one) is sharpness. Evaluated the lens both with a test chart and in the field. My 10 year old 18-55 mm Fujifilm lens consistently comes out on top (I do have a good copy!). Most of the reviewers report opposite results. I have attached a couple of comparison shots; the word "mushy" comes to mind when examining the Tamron shots...have to pixel-peep here at 100% to see the difference. Pictures are unprocessed JPEGs out of camera

FWIW, SOOC JPGs of the two lenses, 55mm at f/10, f/4, and f/2.8. Not very scientific, handheld, tried my best to be consistent. The 17-70 is quite an upgrade over the 18-55 for me, perhaps because I shoot RAW so I didn't notice the problem you're having with SOOC.











and the same 5 shots, exported from RAW in darktable, with no vignetting/distortion correction.

 

[jm10]

I am in the process of evaluating a new Tamron 17-70mm lens. Among other things I noticed that a X-T2 camera does not report correct focal distances between 17 and 24 mm - it is always 17 mm. Starting with 24 mm and beyond it works fine. Has anyone encountered this?

mine also goes from 17mm to 20mm, no 18mm or 19mm

Another problem (a more serious one) is sharpness. Evaluated the lens both with a test chart and in the field. My 10 year old 18-55 mm Fujifilm lens consistently comes out on top (I do have a good copy!). Most of the reviewers report opposite results. I have attached a couple of comparison shots; the word "mushy" comes to mind when examining the Tamron shots...have to pixel-peep here at 100% to see the difference. Pictures are unprocessed JPEGs out of camera

FWIW, SOOC JPGs of the two lenses, 55mm at f/10, f/4, and f/2.8. Not very scientific, handheld, tried my best to be consistent. The 17-70 is quite an upgrade over the 18-55 for me, perhaps because I shoot RAW so I didn't notice the problem you're having with SOOC.

and the same 5 shots, exported from RAW in darktable, with no vignetting/distortion correction.

Thanks sluggy warrior. I appreciate the time you took to make the pictures!

Glad you have the X-T2 So it appears that focal distance mis-reporting is a "feature" for X-T2 <==> 17-70 combination. It was resolved starting with X-T3.

From your pictures I can see that the 17-70 is indeed an upgrade over your 18-55 both for JPEGs and images exported from RAW. In my case the situation is more murky; I have a very good 18-55 copy and while comparing JPEG images with the 17-70 the Fujifilm pictures clearly look better. The processed RAW images look much closer (the resolution), but to my eye the 18-55 lens still looks slightly better.

I am tempted to send the 17-70 back for a replacement.

 

[sluggy_warrior]

Thanks sluggy warrior. I appreciate the time you took to make the pictures!

Glad you have the X-T2 So it appears that focal distance mis-reporting is a "feature" for X-T2 <==> 17-70 combination. It was resolved starting with X-T3.

From your pictures I can see that the 17-70 is indeed an upgrade over your 18-55 both for JPEGs and images exported from RAW. In my case the situation is more murky; I have a very good 18-55 copy and while comparing JPEG images with the 17-70 the Fujifilm pictures clearly look better. The processed RAW images look much closer (the resolution), but to my eye the 18-55 lens still looks slightly better.

I am tempted to send the 17-70 back for a replacement.

I think your 17-70 copy might be underperforming, if it doesn't cost much to send it back, worth a try.

I've just reviewed your photos and mines, and wondering if the X-T2 is having trouble nailing focus with the 17-70. I looked at your EXIF, the focus pixel was on the pine cone yet the actual focus is further than the cone.

Similarly, the focus on my f/4 shot was also slightly behind, but could have been me not holding still and moved forward a bit after acquiring focus.

 

[jm10]

Thanks sluggy warrior. I appreciate the time you took to make the pictures!

Glad you have the X-T2 So it appears that focal distance mis-reporting is a "feature" for X-T2 <==> 17-70 combination. It was resolved starting with X-T3.

From your pictures I can see that the 17-70 is indeed an upgrade over your 18-55 both for JPEGs and images exported from RAW. In my case the situation is more murky; I have a very good 18-55 copy and while comparing JPEG images with the 17-70 the Fujifilm pictures clearly look better. The processed RAW images look much closer (the resolution), but to my eye the 18-55 lens still looks slightly better.

I am tempted to send the 17-70 back for a replacement.

I think your 17-70 copy might be underperforming, if it doesn't cost much to send it back, worth a try.

I've just reviewed your photos and mines, and wondering if the X-T2 is having trouble nailing focus with the 17-70. I looked at your EXIF, the focus pixel was on the pine cone yet the actual focus is further than the cone.

Similarly, the focus on my f/4 shot was also slightly behind, but could have been me not holding still and moved forward a bit after acquiring focus.

I also feel it may be underperforming.

Regarding the pictures with the pine cone: yes, the focus was on the pine cone but in each picture it was on a different section of the pine cone. In retrospect a cleaner experiment would have been to use the same camera, swap lenses, select the smallest focus square and try to focus exactly on the same point. So I am not ready to draw major conclusions from this. But with a f/10 aperture I was expecting to see a greater section of the image in focus.

 

[baobob]

I have been using my XF 18-135mm and XF 18-55mm for years then sold them and replace with the Tamron.

After some thousands shots I just find its IQ similarly to reviewers : excellent even at f 2.8, the AF is fast and altogether for me it is 4.5/5

 

[baobob]

From Lenstip review :

"However, this lens managed to achieve the most important task: despite a very large and very universal focal range combined with a good, fixed f/2.8 aperture fastness, it guarantees a very good image quality across the frame, no matter what focal lenght you employ. It seems the tested Tamron performs even a tad better than you would expect, and, as such, it fully deserves a very high final assessment it got."

 

[baobob]

mine also goes from 17mm to 20mm, no 18mm or 19mm

With the XH2s it is a bit more complicated :

If you set aperture FROM 17mm to a longer FL it records correctly, BUT from longer FL to 17mm in the 17-24mm it is not corect as already mentionned in this thread.

A bit of mistery

 

[sluggy_warrior]

mine also goes from 17mm to 20mm, no 18mm or 19mm

With the XH2s it is a bit more complicated :

If you set aperture FROM 17mm to a longer FL it records correctly, BUT from longer FL to 17mm in the 17-24mm it is not corect as already mentionned in this thread.

A bit of mistery

Do you also have a start-up delay on the XH2s?

On the X-T2 and X-T20, whereas Fuji XF lenses take about a second to get ready (count one-thousand-one), the Tamron takes about two seconds (count one-thousand-one-one-thousand-two). I contacted Tamron and they said it's within specs :-(

And the noisy aperture mechanism, no aperture ring, distortion in live view.

But I seem to have adapted to it just fine so the above don't bother me anymore.

I agreed with the reviews that its strength is the consistent IQ across the frame and focal range. I think Tamron focused the IQ at the wide apertures, f/2.8-5.6, thus, its IQ isn't better at f/8 and smaller. Fortunately, f/2.8-4.0 is where I (and others?) use it the most.

 

[baobob]

yes 2 sec but it's the same with the XFs

 

[jm10]

mine also goes from 17mm to 20mm, no 18mm or 19mm

With the XH2s it is a bit more complicated :

If you set aperture FROM 17mm to a longer FL it records correctly, BUT from longer FL to 17mm in the 17-24mm it is not corect as already mentionned in this thread.

A bit of mistery

Thanks for the information Bob. So it appears from this thread that the focus length mis-reporting is camera model dependent: X-T2 and the H-H2s are affected but X-T3 is OK. I have reported this to Tamron. Their technician will get in touch with Japan. The local US technical capabilities are probably quite limited.

But my main concern is image quality. I hope that this is not camera model dependent...