Tamron Lens Phenomenon

[Canon Rumors Guy]

<strong>Interesting issue with Tamron Lenses</strong>

I received this information a while back, I didn’t post it until there was some kind of response from Tamron.</p>
<p><strong>The following is a direct quote from the folks that found the problem. Tamron’s response is at the bottom.

</strong>i am a member of the german dslr-forum (www.dslr-forum.de), which is probably the biggest german speaking community regarding all DSLR-topics (not only Canon). We observed a phenomenon where the followings lenses won’t use the outer cross-type sensors of EOS 40/50/60/7D cameras right.</p>
<p>The lenses that show this problem are (all Tamron):</p>
<ul>
<li>17-50 f/2.8 VC</li>
<li>60 f/2 Macro</li>
<li>70-200 f/2.8</li>
<li>10-24 f/3.5-4.5</li>
</ul>
<p>To get a bit more into the detail of the problem – on the 40/50/60D cameras the outer cross-type sensors only work as line-type sensors. To be a bit more correct, they actually work in the same direction as they would work on cams like the 30D or like the 400/450/500/550D. When it comes to the 7D the outer sensors are also only working as line-type sensors (but it seems, that the EF 50 1.2L also has problems to use them correct).</p>
<p>The center cross-type sensor does work at least as a single-cross type sensor (infact we don’t now how to test if the center sensor is working as a dual-cross-type sensor like it’s supposed to work with those f/2.8 and better lenses).</p>
<p>Tamron seems to not know about this behaviour so far. Well – at least, they didn’t confirmed the problem and only said, that they will investigate this.</p>
<p>Maybe a small addition – the Tamron 17-50 VC uses the same lens-ID as the EF 35-80 – maybe you’ll have the chance to check whether this Canon-lens performs correctly or not.</p>
<p>The link to the thread is:Â <a href="http://www.dslr-forum.de/showthread.php?t=732360  ">http://www.dslr-forum.de/showthread.php?t=732360</a></p>
<p><strong>Tamron Europe’s Official Response

</strong>Dear Mr. S.,</p>
<p>we were able to reproduce the situation where the Tamron 17-50 VC (B005) and the Tamron 60 Macro (G005) won’t use the Canon EOS 50D cross-type sensors correctly.</p>
<p>We have given this information to Tamron in Japan and we are expect an answer very soon.</p>
<p>Best regards

Tamron Europe GmbH</p>
<p><em>Thanks Joerg</em></p>
<p><strong><span style="color: #ff0000;">c</span>r</strong>

 

[ssbuchanan]

I can confirm that on the 1DMkIV the Tamron 70-200 f/2.8, focusing is terrible, even using the cetnre point exclusively. Much worse than on either my 400D or 40D...

 

[nocojoe]

I can see why they tried to avoid the issue. Wow. That makes those lenses defective on those cameras. I don't know if that is something fixable or not. But that can be a huge blow to Tamron.

 

[jsixpack]

Possibly a stupid question, but how can you tell if one of those cameras is using the sensor as cross or line type?

JSP

 

[neuroanatomist]

Possibly a stupid question, but how can you tell if one of those cameras is using the sensor as cross or line type?

By trying to focus on oriented lines - a sensor that's using (or has) only a horizontal line will not lock focus properly on a horizontal line, since the area of contrast will not cross the sensor. A sensor acts as a one-dimensional input, so in the example below, only #1 could be used to achieve focus. #2-4 are what would happen with a horizontal sensor focusing on a horizontal line (although those examples show a vertical sensor with a vertical line). Cross sensors can focus on both horizontal and vertical features.

 

[Diedro]

@jsixpack: I find that out while "playing" with my new camera (I was/still am a bloody beginner), using a FocusTest Chart (http://FocusTestChart.com). When I first posted the problem on a German-speaking DSLR-forum, folks there believed that it is somehow stupid to have such an approach in getting knowledge about your camera / about photography in general. In the meantime I got two responses from Tamron, but I do not have the impression that they were able so far to solve the problem (at least they are not telling that they have a solution, or rather, they are not yet admitting that they have a problem...)
To be continued.
Diedro
(http://www.dslr-forum.de/showthread.php?t=732360)...

 

[Canon Rumors Guy]

Tamron Lens Phenomenon Update

<p><strong>UPDATE #2

</strong>New Information about that phenomenon, which is not just a Tamron phenomenon anymore. A guy at the DSLR-forum replaced the Lens-IDs with an EF adapter and found out that the cross-type sensors won’t work on Lenses with the following IDs: 27, 30, 35, 37, 38, 193, 194, 195).</p>
<p>So Tamron was just unlucky by taking the wrong IDs.</p>
<p><strong>The following Canon lenses are also affected:</strong></p>
<ul>
<li>27 = Canon EF 35-80mm f/4-5.6 (I)</li>
<li>30 = Canon EF 35-105mm f/4.5-5.6</li>
<li>35 = Canon EF 35-80mm f/4-5.6 II</li>
<li>37 = Canon EF 35-80mm f/4-5.6 III</li>
<li>38 = Canon EF 80-200mm f/4.5-5.6</li>
<li>193 = Canon EF 35-80mm f/4-5.6 USM</li>
<li>194 = Canon EF 80-200mm f/4.5-5.6 USM</li>
<li>195 = Canon EF 35-105mm f/4.5-5.6 USM</li>
</ul>
<p><strong>Link: <a href="http://www.dslr-forum.de/showpost.php?p=7463669&postcount=407">http://www.dslr-forum.de/showpost.php?p=7463669&postcount=407</a></strong></p>
<p><em>thanks Jorg</em></p>
<p><strong>UPDATE</strong>

The Tamron 70-200 f/2.8 is also confirmed to have the issue with the 40D, 50D and 7D outer focus points not being crosstype.<strong> </strong></p>
<p><strong>7D Affected</strong>

A few days ago we <a href="http://www.canonrumors.com/2010/11/tamron-lens-phenomenon/">reported some Tamron lenses weren’t using the outer crosstype</a> sensors on the Canon 50D.</p>
<p>Tamron Europe has confirmed the Canon 7D and Tamron 17-50 f/2.8 VC also exhibit the same issue.</p>
<p>We’re still waiting for an official response from Tamron in Japan.</p>
<p><strong><span style="color: #ff0000;">c</span>r</strong></p>

 

[Catastrophile]

maybe the peripheral points aren't really cross typed or they aren't there at all!!!

 

[scalesusa]

I believe that programmable AF confirm chips can be set to send any lens ID you wish to the camera. I assume that this was what was used to send different ID's to the camera. (A Nikon to EOS or other adapter with programmable AF confirm chip)

 

[Edwin Herdman]

I'm surprised nobody's chimed in with a *cue gong hit* conspiracy theory yet. From the post, there's nothing to prove it one way or another - though to my eyes the list of Canon lenses seems to comprise mostly older, outdated designs, all variable aperture, rather old-fashioned focal lengths mostly. If the numeric ID of a lens determines whether it can use the outer cross-type sensors, that's strong evidence that those IDs are being locked out by design - a decision made at Canon. The conspiracy theory would go - it would be little loss to Canon to lock out these older designs, but a bigger loss to Tamron if their better designs (is that the new 70-200mm f/2.8 in the list?) were locked out. What's not explained is why Tamron would have chosen those IDs in the first place, especially if their IDs were taken from lenses of a completely different type (I don't see any lenses in the original Canon ID list that are a wide-angle zoom, like the 17-50 or the 10-24, no macro either). Perhaps there was some other problem with using an ID taken by another lens of the same type (i.e. registering the same ID for the 70-200mm f/2.8 as a more current Canon 70-200mm f/2.8), possibly conflicting microadjustment saved settings? Or does Canon itself not use the simple lens ID in its newer lenses anymore, and Tamron can't copy the new identifying format?

 

[Osiris30]

I'm surprised nobody's chimed in with a *cue gong hit* conspiracy theory yet. From the post, there's nothing to prove it one way or another - though to my eyes the list of Canon lenses seems to comprise mostly older, outdated designs, all variable aperture, rather old-fashioned focal lengths mostly. If the numeric ID of a lens determines whether it can use the outer cross-type sensors, that's strong evidence that those IDs are being locked out by design - a decision made at Canon. The conspiracy theory would go - it would be little loss to Canon to lock out these older designs, but a bigger loss to Tamron if their better designs (is that the new 70-200mm f/2.8 in the list?) were locked out. What's not explained is why Tamron would have chosen those IDs in the first place, especially if their IDs were taken from lenses of a completely different type (I don't see any lenses in the original Canon ID list that are a wide-angle zoom, like the 17-50 or the 10-24, no macro either). Perhaps there was some other problem with using an ID taken by another lens of the same type (i.e. registering the same ID for the 70-200mm f/2.8 as a more current Canon 70-200mm f/2.8), possibly conflicting microadjustment saved settings? Or does Canon itself not use the simple lens ID in its newer lenses anymore, and Tamron can't copy the new identifying format?

It could also be just as likely the old Canon lenses just don't work right with the outer points in cross mode, and so Canon has chosen to disable them.

 

[Edwin Herdman]

Could be. You could compare those lenses to others of their vintage - the lot of them seems mainly to comprise a few families of zoom lens, though, so it might not be a fair comparison.

 

[unexposure]

I'm surprised nobody's chimed in with a *cue gong hit* conspiracy theory yet. From the post, there's nothing to prove it one way or another - though to my eyes the list of Canon lenses seems to comprise mostly older, outdated designs, all variable aperture, rather old-fashioned focal lengths mostly. If the numeric ID of a lens determines whether it can use the outer cross-type sensors, that's strong evidence that those IDs are being locked out by design - a decision made at Canon. The conspiracy theory would go - it would be little loss to Canon to lock out these older designs, but a bigger loss to Tamron if their better designs (is that the new 70-200mm f/2.8 in the list?) were locked out. What's not explained is why Tamron would have chosen those IDs in the first place, especially if their IDs were taken from lenses of a completely different type (I don't see any lenses in the original Canon ID list that are a wide-angle zoom, like the 17-50 or the 10-24, no macro either). Perhaps there was some other problem with using an ID taken by another lens of the same type (i.e. registering the same ID for the 70-200mm f/2.8 as a more current Canon 70-200mm f/2.8), possibly conflicting microadjustment saved settings? Or does Canon itself not use the simple lens ID in its newer lenses anymore, and Tamron can't copy the new identifying format?

Go, check the release-dates of 60 macro, 17-50vc and 10-24 and compare them to the release-date of 40d. then go and share your conclusion. ;-)

 

[Edwin Herdman]

All newer, of course. So quite obviously that knocks out that potential for "proof." It remains an unanswered question why Canon took this heavy-handed action against that range of IDs however - from 2007 on they certainly must have known other manufacturers might use that range of lens IDs. On the other hand, the Canon lenses targeted all are older zooms - lower performance ones at that, all from three types, all seem pretty similar in a regard. All designed for film and at least a good number look like kit zooms for cheaper film cameras.

Since those Tamron lenses are newer than the 40D (i.e. the 60mm F2 Macro was announced after the release of the last 40D firmware) at least some of the blame appears to lie with them for not testing these thoroughly.