From Bev.Oke@nrc.ca Mon Jan 21 17:00:58 2002 Date: Mon, 22 Oct 2001 14:38:40 -0700 From: "Oke, Bev" To: "'N.G.Douglas@astro.rug.nl'" Cc: "Oke, Bev" Subject: Your paper "The Planetary Nebulae Spectrograph: Galaxy Kinematics using Planetary nebulae" [ The following text is in the "iso-8859-1" character set. ] [ Your display is set for the "US-ASCII" character set. ] [ Some characters may be displayed incorrectly. ] October 22, 2001 Dear Dr. Douglas: Your paper "The Planetary Nebulae Spectrograph: Galaxy Kinematics using Planetary nebulae" was sent to an expert referee who has provided the report below. To save time and avoid postal delays I am sending the remarks electronically and will retain your original manuscript and figures here unless you need them back. I have read through the comments of the referee and find that I am in agreement with the referee on most of them. In view of the comments of the referee this paper needs major changes before becoming acceptable. Specific notes: (1) Appendix A. As suggested by the referee, a figure showing the spot diagrams would be far more useful to general readers than what has been given. (2) The footnotes are all missing on page 1. Also, Email addresses should be included in the addresses. (3) The tables are not in the format which is used by PASP. They must be! When you send a revised manuscript, please outline the revisions you have made in response to the referee's comments in a cover letter. Please acknowledge receipt of this message. Yours sincerely, J. B. Oke ---------------------------------------------------------------------------- Report of Referee: Report on "The planetary nebula spectrograph .... " by Douglas et al. --------------------------------------------------------------------- Summary ------- This gives details of an instrument dedicated to studies of planetary nebulae. It is related to earlier publications by the same principal author which give details of application of this technique to existing double-beam spectrographs at the AAT and WHT. This paper presents results from a new purpose-built instrument. Unfortunately, the discussion of the design and construction is rather sketchy and the results of the characterisation of the instrument are not clearly presented. Since no astrophysical results are presented either, I consider that this paper needs substantial modification - chiefly to provide the missing information - before it is suitable for publication by PASP. The style and layout of the paper also needs a lot of attention. I think there is the germ of a good paper here but it needs a lot more work. General comments ---------------- The motivation for this work is clearly stated but is already covered by Douglas & Taylor (1999) so need not be repeated at length in this paper. Although the principle of the instrument is adequately presented, much necessary detail is ommitted even though it has a large impact on the design. Interesting statements are made and analyses presented, but sometimes without enough supporting information (generally unreferenced) to assess their value. It is possible that some of this information is actually there, but it is lost in the confusing layout of the paper - even the abstract seems to belong to a different paper. For this reason, I am confining myself to general comments rather a detailed line by line analysis in the expecation that his can be done when/if the MS is revised. The paper suffers from a number of typographical errors which make it harder to referee. The most obvious is that the sequence "eq" appears to be missing throughout (probably a global editing error).. but there are others (in equations and presentation of results) which should have beeen avoided. In addition, the tables are given "in-line" rather by by the usual practice of reference adopted by all journals. Some of the figures are unclear. Footnotes to tables are missing. Despite my resolve not to go into too much detail, I ended up making many more comments than usual, so please excuse any typos and misunderstandings. Specific comments ----------------- Abstract: not enough information here and incorrect. The "novel" instrument is not simply "proposed" but demonstrated (although inadequately). Some specific results should be quoted - e.g. the accuracy to which radial velocities can be recovered. Sect 2.1: Equation (1) is useful for what comes after. The remainder is for information only since its results are not used elsewhere and could usefully be ommitted. I suggest that a reference to a recent review of PNe properties is given instead. This section is also very similar to Section 2.1 of Douglas & Taylor. Sect 2.2: I haven't verified each calculation but the results seem generally sensible but some important detail is missing........ Sect 2.2.1, eqn 4 and following: It should be made clear exactly what F_PN is. Is it just the flux from the 5007 line (F_5007 as given by Eqn 1, but converted to a detected photon rate)? The assumed readout noise is not stated. Sect 2.2.1, eqn 5 and following: Why is the multiplier of (nr^2) not A_fibre instead of A_PSF? A_fibre should be given in units of pixels not arcsec^2. Again, what exactly is F_PN in this context? Sect 2.2.2, eqn 5 and following: B/N and T/N should both be in parenthesis to avoid confusion in the equation. ...2nd para on p10: "S/N can hardly be surpassed" - but the integration time is reduced by a factor N comparted with direct imaging....? ...3rd para on p10: what is the exposure time? Sect 2.2.3, eqn 7: ** unbalanced parentheses ** ..... probably should be an open bracket in front of f_back. Sect 3.1, penultimate sentence of top para on p15: "..acceptable l100 pixels.." is garbled - presumably should read "..acceptable 40 pixels.." since the bandwith is 40A and the dispersion is 1A/pixel. Sect 3.1, bottom of p15: Fig. 5 is not acceptably clear in my copy. There is an important point which is not stated here. Fig 4 shows a "grating normal to camera" configuration, but Fig. 5 shows a "grating normal to collimator" configuration. The latter gives lower spectral resolving power for a given ruling density and so may require the use of a larger collimated beam. However it makes the camera smaller and permits the more convenient "concave" grating configuration. I think this should be pointed out since it has a large effect on the optical design. Sect 3.1, last para: "..gave good (0.8306) reflectance..". It should be clearly stated that this is a theoretical result - "gave" is a very imprecise word to use. Maystre's code should either be properly referenced as a publication or, if unpublished, more information given in this paper - what does the code actually do? There appears to be come confusion over which dispersion order is being used here: is it +1 or -1? More detail is needed on the sentence "Stray light appears..." - exactly how was this assessed? Sect 3.2: How exactly was the "receiver operating curve" derived - from numerical simulations, theory or actual data? Either a reference should be supplied (if it's got a fancy name, it is presumably a non-standard, non-trivial technique!) or some more description is required. How are the results shown in Fig. 6 normalised. Even 10 false detections would be acceptable if the _total_ number of detections was 1000! The leftmost curve is not identified either in the text or the caption (only in the graph title which would normally be ommitted). I'm not really sure of the benefit of this sub-section anyhow. It seems to be making the obvious point that if you add an extra beam you are going to lose light and thereby reduce your sensitivity. Sect 4.1 p18: the last two paragraphs give instrument performance results which would be better given after the description on the optical and mechanical design. It looks as if these key numbers (throughput and flexure) appear as a summary of what is to come, but there is no subsequent text to give details of how these results were obtained. This, surely, is the whole point of the paper! Is the pixel size 17um or is this the total flexure? >From the "specification" given later, the pixel size is 13.5um, so presumably the author meant to write "27" not "17". 2 pixels seems to be a lot of flexure - so why do you say it is insignificant? If all this flexure actually occurred over a small range of orientation angle, it might well be very significant because the _rate_ of flexure (image motion/hour) might be locally very large. Furthermore, you have already stated that the instrument has the advantage of only being sensitive to the _differential_ flexure between the two cameras. It's not clear which kind of flexure you are referring to here. Sect 4.2 1st para: It would be helpful to remind the reader that the design shown in Fig 5 is being used. It was introduced earlier in the paper without making it clear that this was not just another prototype design (like Fig. 4).... if this is indeed the case! Sect 4.2 last para: Please show the ray-traced image quality (spot diagrams or enclosed energy). I don't think it is helpful to simply refer to an appendix which itself just refers to some web pages. With Zemax it is trivial to derive these diagrams - or I'm sure Damien Jones could do this. The last two sentences are opaque and unhelpful - please state what the image quality was: (a) from ray-tracing, (b) found in practice. Sect 4.3: please add someting like "see below" after mention of Fig. 9 to stop the reader worrying about what the mask is for. You should also say why it is desirable to tilt the filter? The 2nd paragraph deals with the all-important differential flexure between the two arms. This looks very good but is a theoretical result. How does this relate to the "result" presented at the end of Section 4.1 (17um)? Sect 4.5: "Specifications" are written down _before_ you start building the instrument. This section would be better called something like "instrument parameters" to make it clear that you are listing the as-built capabilities of the instrument. The tables should not be given in-line in the text but referenced as "Table 1" etc. and be provided separately with table captions. What exactly does "without retuning" mean? Some of entries include superscripts which presumably refer to footnotes - but these are not provided! Sect 4.6: Same point about tables as above. The term "calculated" should be replaced by "predicted" or "theoretical". It should be clearly stated that the first table in this section gives the "predicted efficiency of the instrument optics ... excluding the telescope, detector, filter...." ? The next table gives the predicted efficiency of the items not included in the previous table. How was the "bulk absorption" of the glass worked out? - (indeed, which glasses are used in the instrument?) Was the Melles Griot a/r coating mentioned actually used in the real instrument? The detector is stated as "Tek" but what kind of Tektronix detector was used? Section 4.7: This needs more detail and would have been a good place to discuss the measured on-sky performance of the instrument. Sect 4.8, end of 2nd paragraph. Either Ciardullo's results can be referenced to a learned publication or you need to give more details of the work here - a "priv. comm." is not enough. Section 5: We have come to the end of the paper but *** I have seen no clear statement on whether the instrument reached its design goals ***. What was the actual throughput and flexure? Maybe you gave this in Section 4.1 but you did not make it clear how these results were obtained and under what circumstances they are valid..... Appendix A: I will leave it to PASP to say whether they want references to detailed design data to be in this paper. My view is that you should give more information in the text (e.g. spot diagrams) and dispense with this appendix. Appendix B: This is genuinely interesting and useful but I haven't gone over it in fine detail. App B.3: I would have thought that a reasonably efficient 600/mm grism pair could have been found. The author is right to say the VPH gratings would be a natural alternative to consider. References: I haven't gone over this in detail, but obviously the isolated entry of "private communication" is not very helpful. Figures: Fig 1: I don't understand why you couldn't wait until you had this figure ready before submitting the paper. Fig 2: the caption needs to explain the significance of the "Resolution elements" see in the cartoon spectra. Fig 3: the caption needs to make it a little clearer about what is shown in the inset. Fig. 5: Needs to be much improved in clarity. In general, raw engineering drawings should be avoided. Fig 6: caption needs to make it clear which curve is which. What do the dash-dot-dot lines refer to? Fig 7: probably not necessary. Fig 11: see comments about Ciardullo "priv. comm." above.